Determination of Phenolic
Antioxidant DBPC and DBP Levels in
Authors
Dipak Mainali and Frank Higgins
Agilent Technologies, Inc.
Electrical Insulating Oil
Agilent 5500, 4500, and Cary 630 FTIR Spectrometers
Application Note
Introduction
The phenolic antioxidants 2,6-ditertiary-butyl paracresol (DBPC) (also known as
butylated hydroxytoluene (BHT)) and 2,6-ditertiary-butyl phenol (DBP) are the two
most common oxidation inhibitors added to the electrical insulating (transformer) oil
and mineral oil based lubricants. The typical recommended value of DBPC and DBP
in fresh electrical insulating oil is approximately 0.3% by weight. These inhibitors
prevent electrical insulating oil from oxidative degradation, and prolong the life of
the oil. It is essential to maintain the optimum concentration level of inhibitors to
ensure the proper functioning of mineral oil used in transformer units as an insulat-
ing or cooling agent. The depletion rate of the inhibitors in the oil is dependent on
various factors such as the amount of oxygen available, soluble contaminants, cat-
alytic agents, and temperature. Therefore, regular testing of the inhibitors in electri-
cal insulating oil is necessary to ensure its reliable operation in high value assets
such as transformer units.
ASTM 2668 and IEC 60666 are the standard test methods, using infrared spec-
troscopy (IR) technology to monitor the concentrations of DBPC and DBP in electri-
cal insulating oil. These test methods are used to determine if the new electrical
insulating oil meets the specification for oxidation inhibitor initial concentration
levels. They also monitor the concentration of inhibitors in the used oil. If the
inhibitors have been depleted to a critical level, additional inhibitor can be added.
Therefore, the standard test methods are essential for manufacturing control,
specification acceptance, and to periodically monitor the level of inhibitors in
used oils.
This application note describes the methods developed fol-
lowing ASTM 2668 and IEC 60666 to measure the concentra-
tion level of DBPC and DBP in electrical insulating oil. The
methods described are suitable for all mineral oil base stocks
[1], including turbine oils, hydraulic oils, gear oils, compressor
oils, and crankcase oils. The ASTM and IEC methods
described in this application note are able to measure up to
1 wt.% DBPC and 0.8 wt.% DBP concentration in new or used
mineral oil, which is higher than the amount covered by
ASTM 2668 or IEC 60666 test methods, which only cover up
to 0.5 wt.%. The methods are developed using the high per-
forming Agilent 5500, Agilent 4500, or Agilent Cary 630 FTIR
spectrometers, and the measurements using these methods
are quick, easy, and can be performed on-site with the mobile
5500 and fully portable 4500 FTIR.
2
Methods and Materials
To develop the ASTM 2668 method, the calibration samples
were prepared from an uninhibited standard mineral base oil
without phenolic antioxidants. The standard mineral oil was
obtained as Base 20 and Base 76 from SPEX standards. The
calibration samples in the range of 0–1.0% DBPC by weight
were prepared by dissolving DBPC in standard mineral oil
using a high precision analytical balance. The samples were
measured on 5500, 4500, and Cary 630 FTIR spectrometers
with the TumblIR or DialPath transmission cell set at a path-
length of 100 µm (0.1 mm). Each spectrum was collected at
8 cm–1 resolution with 128 co-added scans, yielding the
approximate measurement time of 30 seconds.
To develop the IEC 60666 method, the calibration standards in
the range of 0–0.8% DBPC by weight were prepared by adding
DBPC in uninhibited Base 20 mineral oil obtained from SPEX
standards using a high precision analytical balance. The cali-
bration samples were measured using a TumblIR and a
DialPath transmission cell at three different pathlengths
(200 µm, 500 µm, and 1,000 µm). Each spectrum was collected
at 4 cm–1 resolution with 64 co-added scans.
A B
Figure 1. A) DialPath and (B) TumblIR accessories.
Results and Discussion repeatability and reproducibility of the predicted results much
better than the ASTM-specified 0.04 wt.% limits.
For both ASTM and IEC method calibrations, the region of the
FTIR spectra used to determine the DBPC concentration is the As shown in the validation results in Table 1, the maximum
phenolic OH stretch at 3,650 cm–1 (Figure 2). FTIR spectra variance in this method is 0.01 wt.%, and 0.02 wt.% for the
indicated no interfering differences in the oils with different standards measured in the range of 0–0.5 wt.% and
viscosities in the region of interest for measurement of the 0.5–1 wt.%, respectively. The second measurement range of
phenolic OH stretch at 3,650 cm–1. 0.5–1 wt.% is above the range covered in ASTM 2668 proce-
dure, but still exceeds its performance criteria. Therefore, the
For the ASTM 2668 method, the calibration constants were ASTM 2668 method developed using 5500, 4500, or Cary 630
obtained by performing the linear regression plot of the peak spectrometers covers the measurement of DBPC up to
area (at 3,650 cm–1) with dual baseline against DBPC 1 wt.%, and provides the quantitative measurement of DBPC
concentration, as shown in Figure 3. The calibration curve has with excellent repeatability and reproducibility.
excellent linearity, with R2 = 1.00, and provides the

0.06

0.04

0.02
Absorbance

-0.02

-0.04

-0.06

-0.08
3,705 3,695 3,685 3,675 3,665 3,655 3,645 3,635 3,625 3,615
Wavenumber (cm–1)

Figure 2. The IR spectral overlay of the absorbance band for phenolic OH stretch of DBPC in mineral oil (0–1 wt.%).

Table 1. Predicted Versus Actual Values for

10,000 Calibration plot for DBPC the Calibration Data Set Using the
9,000 R2 = 1.00
ASTM 2668 Method
8,000
ASTM 2668 method
7,000
Concentration (ppm)

DBPC concentration weight%

6,000
Sample no. Actual Predicted1
5,000
1 0 0±0
4,000
2 0.03 0.03 ± 0
3,000
3 0.05 0.05 ± 0
2,000
4 0.1 0.10 ± 0
1,000
5 0.2 0.20 ± 0
0
6 0.4 0.40 ± 0.010
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 7 0.6 0.60 ± 0.016
Peak area
8 0.8 0.80 ± 0.016
Figure 3. The calibration plot for DBPC (aka BHT) in mineral oil base stock 9 1 1.0 ± 0.02
in ppm units, multiply ppm by 0.0001 for weight% values. 1 Average of four values measured in four different
instruments ± two standard deviations

3
Figure 4 shows that, for the IEC 60666 method, a calibration In the Agilent MicroLab methods for both ASTM and IEC, the
curve for the standards measured at a 1,000 µm pathlength user can set threshold limits (Figure 4) as their analysis
was obtained by plotting absorbance at 3,650 cm–1 against demands. The final result is displayed in a color code (red,
the DBPC weight percent values. The calibration has excel- yellow, or green) emphasizing the state of DBPC in the ana-
lent linearity, with R2 = 0.999. The calibration curves looked lyzed oil sample (Figure 5). Similarly, the recommendation
similar for the samples measured at 200 µm and 500 µm based on the threshold limit can be described in the MicroLab
pathlengths. method, which would be displayed at the end result, prompt-
ing the analyst to take an appropriate action (Figure 6). This
0.9 y = 1.0647x + 0.0076 unique color coding and the recommendation feature of
0.8 R2 = 0.9999 MicroLab software makes it easier for a new operator to
0.7 understand the result, and take appropriate action.
DBPC weight %

0.6
0.5
0.4 Conclusion
0.3
0.2
Both ASTM D2668 and IEC 60666 methods developed using
0.1 Agilent 5500, Agilent 4500 FTIR, or Agilent Cary 630 spec-
0 trometers with TumblIR and DialPath transmission cells
0 0.2 0.4 0.6 0.8
Absorbance provide the sensitive results necessary to assist personnel
Figure 4. The calibration plot for DBPC in mineral oil measured at a
monitoring the DBPC and DBP levels in electrical insulating
pathlength of 1,000 µm. oils. The methods are designed to alert the analyst using
preset warning levels when the phenolic antioxidants are at
or approaching depletion limits. This enables the analyst to
Table 2 shows the validation results using the IEC method for maintain the proper level of DBPC or DBP in oil used in high
the samples measured using the 1,000 µm (1 mm) pathlength. value assets such as transformer units, turbines, and engines.
The repeatability and reproducibility were well within the
specified limits of the IEC 60666 procedure. In addition, the ability of 5500 and 4500 FTIR spectrometers to
measure DBPC and DBP on-site eliminates the hassle and
cost of sending the samples to an off-site laboratory. The
Table 2. Actual Versus Predicted Values for
the Calibration Set Using IEC 60666
measurements are rapid and minimize the dependency on the
Method skill of the operator due to the intuitive usability of the
MicroLab software methods.
IEC 60666 method
DBPC concentration weight%
Sample no. Actual Predicted
1 0.00 0.00
2 0.10 0.09
3 0.20 0.19
4 0.03 0.03
5 0.05 0.05
6 0.60 0.60
7 0.80 0.82

4
Figure 5. The MicroLab software feature in which the user can define the threshold limit for the DBPC concentration.

5
Figure 6. The MicroLab final result screen display where the results are shown color coded, with a recommen-
dation. The green color indicates that the DBPC is at the desired level, whereas the red color indicaties
that the DBPC is depleted to the critical level.

Reference
1. F. Higgins, Onsite additive depletion monitoring in turbine
oils by FTIR spectroscopy, Agilent Technologies,
publication number 5990-7801EN (2011).
www.agilent.com/chem
For More Information Agilent shall not be liable for errors contained herein or for incidental or consequential
damages in connection with the furnishing, performance, or use of this material.
These data represent typical results. For more information on
our products and services, visit our Web site at Information, descriptions, and specifications in this publication are subject to change
without notice.
www.agilent.com/chem.
© Agilent Technologies, Inc., 2015
Printed in the USA
November 12, 2015
5991-6380EN